COVID-19 VACCINE MISINFORMATION DETECTION

Abstract

In the situation of pandemic there has been a rise in the act of spreading rumors and myths on online social media platforms and Internet blogs. This has caused harm to the society and has caused people to believe in this fake news. In this study we will be working on classifying the news as fake or real and helping people to stay safe from these myths. We are working on classifying the misinformation related to vaccines. There have been rumors since the development of the vaccine started, news about the deaths during the vaccine trials, cost of vaccine, news regarding shortage of the vaccine has been circulating on various platforms. We have referred a dataset which contains tweets regarding vaccine which has keywords such as vaccine, covid-19, covaxin, sputnik etc. These has been trained using various NLP techniques like BERT, combined BERT with CNN and BiLSTM, RoBERTa, ALBERT. At last, we have compared the results obtained from these models.